- Email: [email protected]
Evidence of metabolic syndrome in lean children with premature pubarche at diagnosis
Available online at www.sciencedirect.com
Metabolism Clinical and Experimental 57 (2008) 733 – 740 www.metabolismjournal.com
Evidence of metabolic syndrome in lean children with premature pubarche at diagnosis Revi P. Mathew a,⁎, Daniel W. Byrne b , MacRae F. Linton c , Douglas E. Vaughan c , Sergio Fazio c , William E. Russell a,d a Department of Pediatrics, Vanderbilt University, Nashville, TN 37232, USA Department of Biostatistics, Vanderbilt University, Nashville, TN 37232, USA c Department of Medicine, Vanderbilt University, Nashville, TN 37232, USA d Department of Cell and Developmental Biology, Vanderbilt University, Nashville, TN 37232, USA Received 12 May 2007; accepted 7 January 2008 b
Abstract We investigated for evidence of early metabolic syndrome irrespective of body mass index (BMI) in subjects with premature pubarche (PP). Ten children with PP were compared with controls matched for age, sex, ethnicity, and BMI. Congenital adrenal hyperplasia and other known causes of PP were excluded by standard methods. Anthropometry, blood pressure (BP), dual-energy x-ray absorptiometry body scan, fasting blood lipid profile, and cytokines were obtained. The children were divided into 2 groups: (1) the total group of children with PP, and their age-, sex-, ethnicity-, and BMI-matched controls and (2) those with PP and normal BMI (b19 kg/m2) and their matched controls selected from the original groups. The PP subjects with normal BMI (S1) showed significantly higher systolic BP (P = .028), diastolic BP (P = .028), and mean arterial pressure (P = .018) compared with matched controls (C1). Nevertheless, for both groups, all the above parameters were statistically not significant when corrected for height. Fat distribution in PP subjects indicated significantly higher android (P = .047) and android-gynoid ratio (P = .013). Normal-BMI PP children had significantly higher android-gynoid ratio fat distribution compared with their matched controls (P = .037). Trunk fat percentage (p: 0.04) and trunk fat (grams) (P = .007) were significantly elevated in PP children compared with matched controls. Again, for both groups, all the above parameters were not statistically significant when corrected for height. The PP subjects had significantly higher tumor necrosis factor (TNF)–α (P = .038) and interleukin-8 (picograms per milliliter) (P = .05) compared with matched controls. Normal-BMI PP children also had higher TNF-α (P = .028) compared with matched controls. When corrected for height, TNF-α was higher in the total (P = .037) and normal-BMI (P = .043) PP children. Premature pubarche can be linked to markers of the metabolic syndrome in lean children. Published by Elsevier Inc.
1. Introduction Premature pubarche (PP), the early appearance of pubic hair before the age of 8 years (9 years in boys) [1], is almost 10 times more common in girls than in boys [2]. Traditionally, PP has been considered a benign process [2]. Adolescent girls with a history of PP, however, have an increased incidence of polycystic ovary syndrome [3], which is usually associated with obesity, hyperinsulinemia, hyperandrogenism, and dyslipidemia [4,5]. The metabolic syndrome (visceral obesity, dyslipidemia, hyperglycemia, and hypertension) has become one of the major public health ⁎ Corresponding author. Tel.: +1 615 936 1874; fax: +1 615 343 5845. E-mail address: [email protected] (R.P. Mathew). 0026-0495/$ – see front matter. Published by Elsevier Inc. doi:10.1016/j.metabol.2008.01.010
challenges worldwide [6].There is no universally accepted or defined criteria for diagnosing metabolic syndrome in children [7]. As a result, different groups, for their studies or analyses, used their own modifications of adult criteria, usually those of Adult Treatment Panel III [8] but, occasionally, a combination of criteria from different definitions [9-12]. The ultimate importance of early diagnosis of metabolic syndrome is that it helps identify individuals at high risk of both type 2 diabetes mellitus and cardiovascular disease (CVD) [13] and should lead to early implementation of prevention strategies including lifestyle changes [14,15]. Hyperinsulinemia and obesity are common features in prepubertal and pubertal girls with a history of PP [5]. In addition, boys with PP show reduced insulin sensitivity
734
R.P. Mathew et al. / Metabolism Clinical and Experimental 57 (2008) 733–740
independent of obesity [16]. Ibanez et al [17] have shown that PP girls had excess total body and central fat throughout all pubertal stages, and increased central fat was related to hyperinsulinemia and hyperandrogenemia. The importance of fat distribution, in addition to total fat, as a risk factor for CVD is recognized in both adults [18] and children [19]. An android or male fat pattern, with relatively greater fat in the upper body region, is associated with negative metabolic predictors [19,20]. A gynoid or female fat pattern, with relatively greater fat in the hip and thigh areas, is associated with less metabolic risk [21]. Similarly, a pattern of fat deposits favoring trunk relative to extremities has also been linked to health risks [22,23]. Differences in body and extremity fat composition have not been previously reported in children with PP and normal body mass index (BMI). Clinical studies in children have demonstrated that an atherogenic pattern of risk factors can start in childhood [24,25]. Such a pattern has been noted in children with PP [24], but that study was limited to girls. Obesity and cardiovascular risk factors track from childhood into adulthood [25,26]. Thompson et al [26] have shown that CVD risk factors are present in overweight children at the age 9 of years, but this has not been demonstrated in lean children with PP. In this study, we demonstrate that both boys and girls with PP have significantly higher total cholesterol (TC) to highdensity lipoprotein (HDL) ratios compared with controls matched for age, sex, ethnicity, and BMI. Furthermore, increased adiposity, serum dehydroepiandrosterone sulfate (DHEAS), and plasma cytokine levels as early markers of metabolic syndrome were found in lean children with PP. These findings are consistent with early onset of metabolic syndrome in children with PP and normal BMI (b19 kg/m2). 2. Methods
stage II or more pubic hair development in either sex, Tanner stage 1 breast in girls, and testicular volume of less than 4 mL in boys) at diagnosis were included in the study. 3.2. Exclusion criteria Before enrollment in the study, all patients with PP underwent an adrenocorticotropic hormone stimulation test to exclude those with late-onset forms of congenital adrenal hyperplasia, and IV leutinizing hormone releasing hormone (LHRH) [28] and radiological investigations, if indicated, to exclude adrenal and ovarian tumors and central precocious puberty. These tests are standard in the evaluation of children with PP and not part of the experimental design for this study. Children with PP secondary to any of the above diagnoses were excluded from the study. 3.3. Final subject group There were 10 PP subjects (ST): 7 girls and 3 boys (mean age ± SD, 7.3 ± 1 years; age range, 5.9-8.98 years). Seven of the total group of 10 children (5 girls and 2 boys [S1]) were separately analyzed because of their normal BMI (b19 kg/m2, ie, b95th percentile at age 7 years) (mean age ± SD, 7.2 ± 1.2 years; age range, 5.9-8.98 years).
4. Controls Prepubertal controls without PP, but matched for age (±2 years), sex, ethnicity, BMI (±2 kg/m2), and maternal history of gestational diabetes were identified for each of the PP subjects. Hence, 3 control groups were matched one to one for each of the subjects: (a) prepubertal girls younger than 6 years if African American, (b) prepubertal girls younger than 8 years if white, and (c) prepubertal boys of either ethnic origin younger than 9 years.
2.1. Subjects: informed consent
4.1. Method of recruitment
All children enrolled in the study and their parents were thoroughly familiarized with the protocols, including any risks of the procedure described below and the potential benefits gained as a result of the study. The study was approved by the Vanderbilt University Institutional Review Board Human Subjects Committee. Participants were permitted to terminate their involvement in the project at any time during their enrollment. Written informed consent was obtained from parents or guardians of all participants, and assent was also obtained from the participant child.
Control children were recruited from the General Pediatric or Adolescent Clinic, local elementary schools, private pediatric practices, or through advertisement in the Vanderbilt University media. 4.2. Exclusion criteria Controls were screened for hyperlipidemia, and those with fasting cholesterol N200 mg/dL and/or triglycerides N150 mg/dL were excluded as controls. 4.3. Final control group
3. Recruitment of subjects 3.1. Eligibility criteria Girls before the age of 6 years, if African American, or before the age of 8 years, if white [27], and boys below the age of 9 years in either ethnicity with isolated PP (Tanner
There were 7 girls and 3 boys in the matched total control group (CT) (mean age ± SD, 7.3 ± 1.6 years; age range, 3.499.25 years), from which 5 girls and 2 boys were selected based on their match to normal-BMI PP group (S1) and called normal-BMI control group (C1) (mean age ± SD, 7.3 ± 2 years; age range, 3.49-9.25 years).
R.P. Mathew et al. / Metabolism Clinical and Experimental 57 (2008) 733–740
735
Table 1 Anthropometric characteristics Parameter Subjects n Age (y) Sex F/M Weight SD Weight percentile Height SD Height percentile Body surface area (m2) Waist-hip ratio BMI (kg/m2) BMI (SD) Systolic BP (mm Hg) Systolic BP corrected for height (mm Hg/m) Diastolic BP (mm Hg) Diastolic BP corrected for height (mm Hg/m) MAP (mm Hg) MAP corrected for height (mm Hg/m)
All a
Normal BMI a
PP (ST) 10 7.3 ± 1.0 7:3 1.2 ± 1.1 78.2 ± 23.8 1.2 ± 0.8 81.9 ± 13.1 1.06 ± 0.2 0.9 ± 0.1 18.8 ± 4.7 1.3 ± 0.8 109.5 ± 9 84.7 ± 8
Controls (CT) 10 7.3 ± 1.6 7:3 0.8 ± 0.9 69.5 ± 19.1 0.6 ± 0.7 61.6 ± 25.8 ⁎ 1.04 ± 0.2 0.9 ± 0.1 19.0 ± 4.8 1.2 ± 0.8 102.0 ± 6.7 80.2 ± 8.8
PP a (S1) 7 7.2 ± 1.2 5:2 0.6 ± 0.7 69.5 ± 23.5 0.9 ± 0.6 76.1 ± 11.2 1 ± 0.1 0.9 ± 0.1 16.3 ± 2 0.9 ± 0.5 110.7 ± 10.3 86.7 ± 7.9
Controls a (C1) 7 7.3 ± 2.0 5:2 0.4 ± 0.3 61.1 ± 13 0.6 ± 0.6 61.3 ± 24 1 ± 0.2 1 ± 0.2 16.4 ± 1.5 0.8 ± 0.4 99.3 ± 2.9 ⁎ 80.6 ± 9.1
59.4 ± 6.3 46 ± 6.3
53.0 ± 5.6 ⁎ 41.8 ± 6.6
61.4 ± 6.5 48 ± 6.2
52.2 ± 1.6 ⁎ 42.5 ± 5.8
79.7 ± 10.4 61.6 ± 8.9
70.9 ± 3.9 ⁎ 55.9 ± 7
83.1 ± 10.8 64.9 ± 8.3
70.7 ± 2.7 ⁎ 57.4 ± 6.8
a Continuous variables are presented as means ± SD. MAP indicates mean arterial pressure. ⁎ P ≤ .05.
All subjects and controls underwent initial clinical evaluation by the principal investigator including pubertal staging according to Marshall and Tanner [29] to confirm that both groups were prepubertal.
based on the method of Teitz [33] (Alpha Wasserman, ACE HDL-C). Atherogenic index was calculated from the following formula: (TC − HDL)/HDL [34]. Adiponectin and C-reactive protein (CRP) levels were determined by a double antibody sandwich enzyme-linked immunosorbent assay method.
4.5. Height and weight measurements
6.1. Dehydroepiandrosterone sulfate
Height was measured 3 times using a Harpenden stadiometer, and the average was taken in deciding the actual height. Weight was measured without shoes using a digital weighing machine once. Weight and height SD scores, BMI [30], surface area, and waist-hip ratio were calculated.
The measurement of DHEAS was accomplished using a kit from Diagnostic Systems Laboratories, Webster, TX, according to the manufacturer's instructions. The sensitivity of the assay was around 1.7 μg/dL, and between-assay coefficient of variation was 10%. There is a 40% crossreactivity with dehydroepiandrosterone.
5. Blood pressure measurements
6.2. Cytokines
Blood pressure (BP) in the recumbent position was determined 3 separate times over a span of 15 minutes using the Dinamap (Critikon Corp, Tampa, FL) and once manually by the nurse coordinator: the average of the above 4 readings was recorded for comparison of systolic, diastolic, and mean arterial BP between PP subjects and their matched controls.
Tumor necrosis factor (TNF)–α, interleukin (IL)-8, IL1B, IL-10, IL-12p70, and IL-6 were assayed by human inflammation cytokine cytometric bead array kit (BD Biosciences Pharmingen, San Diego, CA).
6. Measurement of plasma lipids, adiponectin, and C-reactive protein
The dual-energy x-ray absorptiometry (DEXA) scanner used was a Lunar Prodigy model (software version 8.1 GE Lunar Prodigo Pro, MA). The raw scan data, which include values of tissue and bone, were captured and sent to a computer for quantitative assessment. The term android is used for central obesity, and gynoid is used for obesity of the lower body often seen in women [35]. The radiation exposure for the DEXA was 7 mrem.
4.4. Initial evaluation of the phenotype
Fasting blood samples were collected for the measurements. Determination of cholesterol was based on a modification of the method published by Allain et al [31], that of triglyceride was based on the method of Fossati and Prencipe [32], and that of HDL cholesterol (HDL-C) was
7. Dual-energy x-ray absorptiometry scans
736
R.P. Mathew et al. / Metabolism Clinical and Experimental 57 (2008) 733–740
8. Correction of data for height To control for height, systolic, diastolic, and mean arterial pressures were recalculated for individual patient height (meters) [36]. In addition, serum lipids, DHEAS, biometric measurements, and biochemical markers (cytokines, CRP, and adiponectin) were corrected for height (meters); and significant corrected-for-height values (P ≤ .05) were included in the tables. 8.1. Statistics We assessed pairwise differences between PP subjects and controls using the McNemar test for categorical variables, and a paired t test for continuous variables, when normally distributed. Wilcoxon signed rank test was used when the data were skewed. P values less than .05 or less were considered statistically significant. All tests were 2-tailed. Statistical analyses were performed on a personal computer with the statistical package SPSS for Windows (Version 11.5; SPSS, Chicago, IL) and the statistical language R (www.r-project.org ). 9. Results 9.1. Clinical findings Despite matched BMI in subjects and controls, PP subjects (ST) had a significantly higher height percentile (P = .046) and mean arterial pressure (P = .037) compared with matched controls (CT). The PP subjects with normal BMI (S1) also showed significantly higher systolic BP (P = .028), diastolic BP (P = .028), and mean arterial pressure (P = .018) compared with matched controls (C1) (Table 1). Nevertheless, for both groups, despite the trends, all the above parameters were not significant when corrected for height (ST: mean arterial pressure, 0.13; CT:
Fig. 1. The TC/HDL ratio in controls and subjects.
systolic BP, P = .177; diastolic BP, P = .15; mean arterial pressure, 0.123). 9.2. Biochemical data The PP subjects in either group (ST and S1) had significantly higher DHEAS compared with their respective matched controls (CT and C1, respectively) (P = .043) even after correction for height in both groups (ST vs CT total group, P = .021; ST vs CT normal-BMI group, P = .028) (Table 2), confirming our previously described adrenal source of the androgens [28]. The PP subjects also had higher TC/HDL ratio (p: 0.042) (Fig. 1) and atherogenic index (P = .0.042). The TC/HDL ratio (P = .059) and atherogenic index (p: 0.059) in PP subjects were not significantly elevated when corrected for height. There were nonsignificant trends in increased plasma TC, low-density lipoprotein (LDL), and triglycerides (TG) in PP subjects; and the LDL cholesterol (LDL-C) (milligrams
Table 2 Serum lipid and DHEAS characteristics Parameter Subject n TC (mg/dL) b HDL-C (mg/dL) b LDL-C (mg/dL) b LDL/HDL ratio TC/HDL ratio Atherogenic index TG (mg/dL) c TG/HDL ratio DHEAS (μg/dL) d DHEAS (μg/dL)/height (meter) a
All a
PP (ST) 10 158.7 ± 31.2 47.2 ± 11 94.5 ± 22.6 2.1 ± 0.6 3.53 ± 0.8 2.5 ± 0.8 85.2 ± 44.8 1.69 ± 1.04 80.8 ± 39.6 61.6 ± 31.2
Normal BMI a
Controls (CT) 10 140.1 ± 14.3 46.4 ± 8.3 82.6 ± 10.6 1.8 ± 0.3 3.1 ± 0.4 ⁎ 2.1 ± 0.4 ⁎ 55.9 ± 20.5 1.42 ± 0.61 33.9 ± 21.1 ⁎ 24.9 ± 14.5 ⁎
Continuous variables are presented as means ± SD. Cholesterol in milligrams per deciliter × 0.0259 = millimoles per liter. c Triglycerides in milligrams per deciliter × 0.01 = grams per liter. d Dehydroepiandrosterone sulfate in micrograms per deciliter × 0.026 = micromoles per liter. ⁎ P ≤ .05. b
PP a (S1) 7 147.1 ± 25.7 50.6 ± 11.4 84.7 ± 14.9 2.2 ± 0.5 3.0 ± 0.4 2.0 ± 0.4 62.6 ± 22.7 1.83 ± 1.21 74.8 ± 46.9 57.9 ± 38.0
Controls a (C1) 7 141.1 ± 16.7 49.6 ± 8.9 82.6 ± 9.7 1.9 ± 0.3 2.9 ± 0.3 1.9 ± 0.3 46.1 ± 7.7 1.46 ± 0.54 29.3 ± 16.2 ⁎ 22.0 ± 11.0 ⁎
R.P. Mathew et al. / Metabolism Clinical and Experimental 57 (2008) 733–740
737
Table 3 Variations in extremity and trunk fat and fat distribution by DEXA, and abdominal skin fold thickness Parameter Subject N Left arm fat (g) Total left-sided tissue fat (g) Right arm fat (g) Total right-sided tissue fat (g) Arms fat (g) Fat distribution: android Fat distribution: A/G Trunk fat % Trunk fat (g) Total fat (g) Total body fat distribution % Abdominal skin fold (mm)
All a
PP (ST) 10 518.5 ± 408.5 5808.1 ± 4383.6 509.8 ± 421.7 5720.4 ± 4301.7 1028.6 ± 827.4 37.5 ± 15.2 0.88 ± 0.2 30.6 ± 14.1 5635.7 ± 4914.3 11528.6 ± 8684.4 31.3 ± 12.8 21.6 ± 13.2
Normal BMI a
PP a(S1) 7 283.4 ± 142.5 3334.6 ± 1466.8 272.7 ± 145.6 3292.3 ± 1394.8 556.6 ± 287.2 30.2 ± 11.8 0.8 ± 0.2 23.6 ± 9.8 2850 ± 1307.9 6627 ± 2859.1 25.1 ± 9.1 17.2 ± 10.4
Controls (CT) 10 443.8 ± 383.3 ⁎ 4879.1 ± 4033.8 ⁎ 406.6 ± 430 4681.2 ± 3822.2 ⁎⁎ 904.6 ± 814.7 ⁎ 31.6 ± 15.4 ⁎ 0.76 ± 0.2 ⁎ 25.8 ± 13.9 ⁎ 4158.9 ± 3847.1 ⁎⁎ 9560.4 ± 7855.4 27.5 ± 11.9 14. 9 ± 9.2
Controls a (C1) 7 222.4 ± 82.6 2623.4 ± 779.2 218.9 ± 96.7 2538.1 ± 785.4 441.4 ± 177.7 22.6 ± 5.7 0.65 ± 0.2 ⁎ 17.8 ± 4.8 2014.4 ± 810.1 5161.7 ± 1563.2 20.6 ± 3.5 10.2 ± 3.5
A/G indicates android:gynoid ratio. a Continuous variables are presented as means ± SD. ⁎ P ≤ .05. ⁎⁎ P ≤ .01.
per deciliter) was 94.5 ± 22.6 in PP subjects vs 82.6 ± 10.6 in controls (P = .123) (Table 2). 9.3. Biometric parameters The PP subjects had higher fat (grams) in arms (P = .023), total left side fat tissue (grams) (P = .011), and total right side fat tissue (grams) (P = .004). Fat distribution in PP subjects indicated significantly higher android (P = .047) and android-gynoid ratio (P = .013). Normal-BMI PP children had significantly higher android-gynoid ratio fat distribution compared with their matched controls (P = .037). Trunk fat percentage (P = 0.04) and trunk fat (grams) (P = .007) were significantly elevated in PP children compared with matched controls (Table 3). None of the biometric parameters were significantly elevated when corrected for height (fat [grams] in arms, P = .177; total left side fat tissue [grams], P = .139; total right side fat tissue [grams], P = .459; android fat distribution, P = .207; android-gynoid ratio, P = .175; trunk
fat percentage, P = .193; and trunk fat [grams], P = .328) in PP children compared with matched controls in the total and normal-BMI groups. 9.4. Cytokines The PP subjects had significantly higher TNF-α (P = .038) (Fig. 2) and IL-8 (picograms per milliliter) (P = .05) cytokines compared with matched controls. Normal-BMI PP children also had higher TNF-α (P = .028) (Fig. 2) compared with matched controls. The TNF-α was significantly higher in PP subjects compared with controls in both total and normal-BMI groups when corrected for height (total group, P = .037; normal-BMI group, P = .043) (Table 4). The IL-8 (picograms per milliliter) levels in PP subjects were not statistically different when corrected for height (P = .546). Table 4 Serum cytokines (picograms per milliliter), CRP (nanograms per milliliter), and adiponectin (nanograms per milliliter) Parameter
a
Normal BMI
Subject
PP (ST)
Controls (CT)
PP a (S1)
n TNF-α TNF-α/ height (m) IL-8 IL-1B IL-10 IL-12p70 IL-6 CRP
10 1.8 ± 0.9 1.5 ± 0.6
10 0.8 ± 0.7 ⁎ 0.6 ± 0.5 ⁎
7 1.8 ± 1 1.4 ± 0.7
Controls a (C1) 7 0.6 ± 0.2 ⁎ 0.5 ± 0.5 ⁎
2.6 ± 0.4 5.6 ± 1.2 2.7 ± 1.5 1.2 ± 0.2 2.8 ± 0.7 830.7 ± 923.3 14012.7 ± 8165.8
1.7 ± 1.3 ⁎ 3.8 ± 2.4 2.2 ± 1.6 1 ± 0.8 3.0 ± 1.2 1332.7 ± 1529.8 14068.97 ± 8339.2
2.6 ± 0.4 5.6 ± 1.3 2.5 ± 1.8 1.2 ± 0.2 2.8 ± 0.8 640.5 ± 923.4 16256.5 ± 8965.6
1.4 ± 0.5 3.8 ± 2.9 1.8 ± 1.7 0.9 ± 1 2.8 ± 1.4 1332.7 ± 1529.8 17329.9 ± 7567.1
Adiponectin
Fig. 2. Differences in cytokine TNF-α in controls and subjects.
All a
a Continuous variables are presented as means ± SD. ⁎ P ≤ .05.
738
R.P. Mathew et al. / Metabolism Clinical and Experimental 57 (2008) 733–740
10. Discussion After having excluded tumors and central or peripheral activity of the gonads, we confirmed in the study that the principal cause of PP was adrenal hyperandrogenism, as implied in the term premature adrenarche (PA). We found that PP subjects had a significantly higher ratio of cholesterol to HDL; greater peripheral, truncal, and total body fat; and higher plasma levels of cytokines TNF-α and IL-8 than BMImatched controls. Of these, only TNF-α was significantly and consistently elevated in total and normal-BMI groups of PP children when compared with matched controls and when corrected for height in both groups. These changes in serum cytokines in PP children with normal BMI have not been previously reported. Premature adrenarche was previously considered to be a benign condition; but more recently, it has been associated with hyperinsulinemia, dyslipidemia, and hypertension [37]. Elevated serum triglycerides and an increased ratio of LDLC to HDL-C have been found in girls with PA [24]. The dyslipidemia of insulin resistance has been proposed to be a precursor to the development of atherosclerosis [38]. Ibanez et al [24] showed that girls with PA have higher triglyceride, cholesterol, LDL-C, and VLDL levels than their healthy peers through the prepubertal and pubertal periods. Hyperinsulinemia and insulin resistance, both independently and together, have been hypothesized to result in dyslipidemia in subjects with both normal and impaired glucose tolerance [39,40]. Hyperinsulinemia enhances hepatic VLDL synthesis [38] and could directly contribute to the increased triglyceride and cholesterol values in these subjects. Laws and Reaven [41] showed that a high triglyceride and a low HDL-C concentration are strong indicators of insulin resistance [42]. In our study, TC/HDL and atherogenic index were significantly elevated in the total group of PP subjects. Nevertheless, the above findings for both groups did not reach significance when corrected for their respective heights. Guven et al [34], in a study of Turkish girls, demonstrated that those with PA have higher BMI, elevated LDL-C and VLDL-C levels, hyperinsulinemia, and higher BP than their healthy peers. However, cholesterol differences have not been previously demonstrated in a mixed group of children with PP matched with controls for BMI, in addition to age, ethnicity, and sex. There have been major advances in the noninvasive assessment of body composition. In this study, we used supine DEXA to assess body composition [43]. Along with underwater weighing and isotope dilution, DEXA is considered the most accurate method for measurement of body composition [44]. To provide consistency in body composition measurements, all PP subjects and controls in our study underwent DEXA scan measurement of body composition in supine position. We have shown higher android-gynoid ratio fat distribution in PP children with normal BMI compared with matched controls. This is likely
to be directly related to increased DHEAS levels in PP. Sex steroids play important roles in determining body fat mass and its distribution [45] including body fat patterning [46-48]. It has also been suggested that intraadipose sex steroid metabolism is also a determinant of android vs gynoid patterns of body fat [45]. These differences in body and extremity fat composition in children with PP and normal BMI indicate the influence of fat metabolism in PP subjects even when lipid profiles and BMI are in the reference range and comparable with matched controls. Adipose tissues secrete a variety of molecules and adipocytokines, including TNF-α and IL-8 [49]. Previous studies have shown that elevated serum levels of CRP, IL-6, and TNF-α precede type 2 diabetes mellitus [50] and coronary heart disease [51]. In our study, we found marked elevations in TNF-α and IL-8 initially and elevation of TNFα alone for both groups when corrected for their individual height. Elevation of TNF-α in PP children regardless of BMI suggests a possible direct correlation to later development of essential hypertension in metabolic syndrome. Further studies are needed to confirm this hypothesis. In addition, TNF-α was consistently elevated in obese and nonobese children with PP and, if confirmed in bigger studies, may be an early predictor of metabolic syndrome at PP diagnosis. In children with PP, our results support the contention that the cluster of risk factors for CVD starts in childhood [26,52]. The increased insulin concentrations concurrent with altered lipid profile may initiate this process in the prepubertal period [53]. Based on our data, the development of PP regardless of BMI should be interpreted as indicative of insulin resistance, a proinflammatory environment, and a risk for development of metabolic syndrome. Metformin has been advocated for use in children with insulin-resistant syndromes and is considered safe [54,55]. Tumor necrosis factor–α down-regulation may be accomplished by an antibody to TNF-α (as infliximab in rheumatoid arthritis) [56], although the sole use of TNF-α antibody failed to reverse insulin resistance in humans. Our data suggest that additional therapeutic targets may be identified for the inflammatory cytokines that are elevated in children with PP. In conclusion, children with PP, even those with normal BMI, show early signs of the metabolic syndrome and should be targeted with early lifestyle interventions. Our study suggests that TNF-α levels are markers of metabolic syndrome at PP diagnosis. Acknowledgment The study was supported by Independent Clinical Research Grant from Pfizer and by the General Clinical Research Center (GCRC), Vanderbilt University, Nashville, TN, and GCRC Immunology Core through GCRC Grant MO1 RR00095 from the National Center for Research Resources/National Institutes of Health. We are indebted to Dr Marta Hernanz-Schulman for her critical review of the manuscript.
R.P. Mathew et al. / Metabolism Clinical and Experimental 57 (2008) 733–740
References [1] Saenger P, Reiter E. Premature adrenarche: a normal variant of puberty. J Clin Endocrinol Metab 1992;74:236-8. [2] Reiter E. Premature adrenarche. Endocrinologist 1997;7:85. [3] Ibanez L, de Zegher F, Potau N. Premature pubarche, ovarian hyperinsulinism, hyperandrogenism and the polycystic ovary syndrome: from a complex constellation to a simple sequence of prenatal onset. J Endocrinol Invest 1998;21:558-66. [4] Rittmaster R, Deshwal N, Lehman L. The role of adrenal hyperandrogenism, insulin resistance, and obesity in the pathogenesis of polycystic ovary syndrome. J Clin Endocrinol Metab 1993;76:1295-9. [5] DiMartino-Nardi J. Premature adrenarche: findings in prepubertal African American and Caribbean-Hispanic girls. Acta Paediatr Suppl 1999;433:67-72. [6] Eckel R, Grundy S, Zimmet P. The metabolic syndrome. Lancet 2005; 365:1415-28. [7] Jones K. The dilemma of the metabolic syndrome in children and adolescents: disease or distraction? Pediatr Diabetes 2006;7:311-21. [8] Adults., E.P.o.D.E.a.T.o.H.B.C.i., Executive summary of the third report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). JAMA 2001;285: 2486-97. [9] Cook S, et al. Prevalence of a metabolic syndrome phenotype in adolescents: findings from the third national health and nutrition examination survey. 1988-1994. Arch Pediatr Adolesc Med 2003;157: 821-7. [10] De Ferranti S, et al. Prevalence of metabolic syndrome in American adolescents. Circulation 2004;110:2494-7. [11] Goodman E, et al. Contrasting prevalence of and demographic disparities in the World Health Organization and National Cholesterol Education Program Adult Treatment Panel III definitions of metabolic syndrome among adolescents. J Pediatr 2004;145:445-51. [12] Weiss R, et al. Obesity and metabolic syndrome in children and adolescents. N Engl J Med 2004;350:2362-74. [13] Alberti K, Zimmet P, Shaw J. The metabolic syndrome—a new worldwide definition. Lancet 2005;366:1059-62. [14] Grundy S. Metabolic syndrome: a multiplex of cardiovascular risk factor. J Clin Endocrinol Metab 2007;92(2):399-404. [15] Ferrannini E. Metabolic syndrome: solution in search of a problem. J Clin Endocrinol Metab 2007;92(2):396-8. [16] Denburg M, et al. Insulin sensitivity and the insulin-like growth factor system in prepubertal boys with premature pubarche. J Clin Endocrinol Metab 2002;87:5604-9. [17] Ibanez L, et al. Fat distribution in non-obese girls with and without precocious pubarche: central adiposity related to insulinaemia and androgenaemia from prepuberty to postmenarche. Clin Endocrinol 2003;58:372-9. [18] Bouchard C, Bray GA, Hubbard VS. Basic and clinical aspects of regional fat distribution. Am J Clin Nutr 1990;52:946-50. [19] Daniels S, et al. Association of body fat distribution and cardiovascular risk factors in children and adolescents. Circulation 1999;99:541-5. [20] Walton C, et al. Body fat distribution, rather than overall adiposity, influences serum lipids and lipoproteins in healthy men independent of age. Am J Med 1995;99:459-64. [21] Ashwell M. Obesity in men and women. Int J Obes Relat Metab Disord 1994;18:S1-S7. [22] Smoak C, et al. Relation of obesity to clustering of cardiovascular disease risk factors in children and young adults. The Bogalusa Heart Study. Am J Epidemiol 1987;125:364-72. [23] Sardinha L, et al. Subcutaneous central fat is associated with cardiovascular risk factors in men independently of total fatness and fitness. Metabolism 2000;49:1379-85. [24] Ibanez L, et al. Hyperinsulinemia, dyslipidemia and cardiovascular risk in girls with a history of premature pubarche. Diabetologica 1998;41: 1057-63.
739
[25] Freedman D, et al. Inter-relationships among childhood BMI, childhood height, and adult obesity: the Bogalusa Heart Study. Int J Obes 2004;28:10-6. [26] Thompson D, et al. Childhood overweight and cardiovascular disease risk factors: the National Heart, Lung, Blood Institute Growth and Health Study. J Pediatr 2007;150:18-25. [27] Herman-Giddens M, et al. Secondary sexual characteristics and menses in young girls seen in office practice: a study from the pediatric research in office settings network. Pediatrics 1997;99:505-12. [28] Mathew R, et al. Premature pubarche in girls is associated with functional adrenal, but not ovarian, hyperandrogenism. J Pediatr 2002; 102:91-8. [29] Marshall W, Tanner J. Variations in pattern of pubertal changes in girls. Arch Dis Child 1969;44:291-303. [30] Grinstein G, et al. The relationship between birth weight (BW), body mass index (BMI) and insulin sensitivity (SI) in prepubertal Caribbean Hispanic (CH) and black African American (BAA) girls with premature pubarche.Pediatr Res 1999;45:89 [abstract]. [31] Allain C, et al. Enzymatic determination of total serum cholesterol. Clin Chem 1974;20:470-5. [32] Fossati P, Prencipe L. Serum triglycerides determined calorimetrically with an enzyme that produces hydrogen peroxide. Clin Chem 1982;28: 2077-80. [33] Teitz N. Clincal guide to laboratory tests. Vol. 3rd ed. Philadelphia, PA: Saunders; 1995. [34] Guven A, Cinaz P, Bideci A. Is premature pubarche a risk factor of atherogenesis? Pediatr Int 2005;47:20-5. [35] Flier J, Foster D. Eating disorders: obesity, anorexia nervosa, and bulimia nervosa in William textbook of endocrinology. In: Flier J, editor. Eating disorders: obesity, anorexia nervosa, and bulimia nervosa. 9th ed. Philadelphia, PA: Saunders; 1998. [36] Michels K, Greenland S, Rosner B. Does body mass index adequately capture the relation of body composition and body size to health outcomes? Am J Epidemiol 1998;147:167-72. [37] Saenger P, DiMartino-Nardi J. Premature adrenarche. J Endocrinol Invest 2001;24:724-33. [38] Orchard T, et al. Plasma insulin and lipoprotein concentrations: and atherogenic association. J Epidemiol 1983;118:326-37. [39] Abbott W, Lillioja S, Young A. Relationship between plasma lipoprotein concentrations and insulin action in obese hyperinsulinemic population. Diabetes 1987;36:897-904. [40] Laakso M, Sarlund H, Mykkanen L. Insulin resistance is associated with lipid and lipoprotein abnormalities in subjects with varying degrees of glucose intolerance. Atherosclerosis 1990; 10:223-31. [41] Laws A, Reaven G. Evidence for an independent relationship between insulin resistance and fasting plasma HDL-cholesterol, triglyceride and insulin concentrations. J Inern Med 1992;231:25-30. [42] McLaughlin T, et al. Use of metabolic markers to identify overweight individuals who are insulin resistant. Ann Int Med 2003;139:802-9. [43] Lambrinoudaki I, et al. Body composition assessment by dual-energy X-ray absorptiometry: comparison of prone and supine measurements. Metabolism 1998;47:1379-82. [44] Newton R, et al. Percent body fat measured by BIA and DEXA in obese, African-American adolescent girls. Int J of Obesity 2005;29: 594-602. [45] Wake D, et al. Intra-adipose sex steroid metabolism and body fat distribution in idiopathic human obesity. Clin Endocrinol 2007;66: 440-6. [46] Marin P, Bjorntorp P. Endocrine-metabolic patterns and adipose tissue distribution. Horm Res 39:81-85 1993;39:81-5. [47] Rosenbaum M, Leibel R. Role of gonadal steroids in the sexual dimorphisms in body composition and circulating concentrations of leptin. J Clin Endocrinol Metab 1999;77:1041-5. [48] Klein K, et al. Estrogen levels in childhood determined by an ultrasensitive recombinant cell bioassay. J Clin Invest 1994;94: 2475-80.
740
R.P. Mathew et al. / Metabolism Clinical and Experimental 57 (2008) 733–740
[49] Salmenniemi U, et al. Multiple abnormalities in glucose and energy metabolism and coordinated changes in levels of adiponectin, cytokines, and adhesion molecules in subjects with metabolic syndrome. Circulation 2004;110:3842-8. [50] Schimidt M, et al. Markers of inflammation and prediction of diabetes mellitus in adults (Atherosclerosis Risk in Communities Study): a cohort study. Lancet 1999;353:1649-52. [51] Rifai N, Ridker P. Inflammatory markers and coronary heart disease. Curr Opin Lipidol 2002;13:383-9. [52] Arslanian S, Suprasongsin C. Insulin sensitivity, lipids, and body composition in childhood. Is “syndrome X” present? J Clin Endocrinol Metab 1996;81:1058-62.
[53] Ronnemaa T, et al. Serum insulin and other cardiovascular risk indicators in children, adolescents and young adults. Ann Med 1991; 23:63-72. [54] Freemark M, Bursey D. The effect of metformin on body mass index and glucose tolerance in obese adolescents with fasting hyperinsulinemia and a family history of type 2 diabetes. Pediatrics 2001;107:E55. [55] Kay J, et al. Beneficial effects of metformin in normoglycemic morbidly obese adolescents. Metabolism 2001;50:1457-61. [56] Bondeson J, Maini R. Tumor necrosis factor as a therapeutic target in rheumatoid arthritis and other chronic inflammatory diseases: the clinical experience with infliximab (Remicade). Int J Clin Pract 2001; 55:211-6.
Metabolism Clinical and Experimental 57 (2008) 733 – 740 www.metabolismjournal.com
Evidence of metabolic syndrome in lean children with premature pubarche at diagnosis Revi P. Mathew a,⁎, Daniel W. Byrne b , MacRae F. Linton c , Douglas E. Vaughan c , Sergio Fazio c , William E. Russell a,d a Department of Pediatrics, Vanderbilt University, Nashville, TN 37232, USA Department of Biostatistics, Vanderbilt University, Nashville, TN 37232, USA c Department of Medicine, Vanderbilt University, Nashville, TN 37232, USA d Department of Cell and Developmental Biology, Vanderbilt University, Nashville, TN 37232, USA Received 12 May 2007; accepted 7 January 2008 b
Abstract We investigated for evidence of early metabolic syndrome irrespective of body mass index (BMI) in subjects with premature pubarche (PP). Ten children with PP were compared with controls matched for age, sex, ethnicity, and BMI. Congenital adrenal hyperplasia and other known causes of PP were excluded by standard methods. Anthropometry, blood pressure (BP), dual-energy x-ray absorptiometry body scan, fasting blood lipid profile, and cytokines were obtained. The children were divided into 2 groups: (1) the total group of children with PP, and their age-, sex-, ethnicity-, and BMI-matched controls and (2) those with PP and normal BMI (b19 kg/m2) and their matched controls selected from the original groups. The PP subjects with normal BMI (S1) showed significantly higher systolic BP (P = .028), diastolic BP (P = .028), and mean arterial pressure (P = .018) compared with matched controls (C1). Nevertheless, for both groups, all the above parameters were statistically not significant when corrected for height. Fat distribution in PP subjects indicated significantly higher android (P = .047) and android-gynoid ratio (P = .013). Normal-BMI PP children had significantly higher android-gynoid ratio fat distribution compared with their matched controls (P = .037). Trunk fat percentage (p: 0.04) and trunk fat (grams) (P = .007) were significantly elevated in PP children compared with matched controls. Again, for both groups, all the above parameters were not statistically significant when corrected for height. The PP subjects had significantly higher tumor necrosis factor (TNF)–α (P = .038) and interleukin-8 (picograms per milliliter) (P = .05) compared with matched controls. Normal-BMI PP children also had higher TNF-α (P = .028) compared with matched controls. When corrected for height, TNF-α was higher in the total (P = .037) and normal-BMI (P = .043) PP children. Premature pubarche can be linked to markers of the metabolic syndrome in lean children. Published by Elsevier Inc.
1. Introduction Premature pubarche (PP), the early appearance of pubic hair before the age of 8 years (9 years in boys) [1], is almost 10 times more common in girls than in boys [2]. Traditionally, PP has been considered a benign process [2]. Adolescent girls with a history of PP, however, have an increased incidence of polycystic ovary syndrome [3], which is usually associated with obesity, hyperinsulinemia, hyperandrogenism, and dyslipidemia [4,5]. The metabolic syndrome (visceral obesity, dyslipidemia, hyperglycemia, and hypertension) has become one of the major public health ⁎ Corresponding author. Tel.: +1 615 936 1874; fax: +1 615 343 5845. E-mail address: [email protected] (R.P. Mathew). 0026-0495/$ – see front matter. Published by Elsevier Inc. doi:10.1016/j.metabol.2008.01.010
challenges worldwide [6].There is no universally accepted or defined criteria for diagnosing metabolic syndrome in children [7]. As a result, different groups, for their studies or analyses, used their own modifications of adult criteria, usually those of Adult Treatment Panel III [8] but, occasionally, a combination of criteria from different definitions [9-12]. The ultimate importance of early diagnosis of metabolic syndrome is that it helps identify individuals at high risk of both type 2 diabetes mellitus and cardiovascular disease (CVD) [13] and should lead to early implementation of prevention strategies including lifestyle changes [14,15]. Hyperinsulinemia and obesity are common features in prepubertal and pubertal girls with a history of PP [5]. In addition, boys with PP show reduced insulin sensitivity
734
R.P. Mathew et al. / Metabolism Clinical and Experimental 57 (2008) 733–740
independent of obesity [16]. Ibanez et al [17] have shown that PP girls had excess total body and central fat throughout all pubertal stages, and increased central fat was related to hyperinsulinemia and hyperandrogenemia. The importance of fat distribution, in addition to total fat, as a risk factor for CVD is recognized in both adults [18] and children [19]. An android or male fat pattern, with relatively greater fat in the upper body region, is associated with negative metabolic predictors [19,20]. A gynoid or female fat pattern, with relatively greater fat in the hip and thigh areas, is associated with less metabolic risk [21]. Similarly, a pattern of fat deposits favoring trunk relative to extremities has also been linked to health risks [22,23]. Differences in body and extremity fat composition have not been previously reported in children with PP and normal body mass index (BMI). Clinical studies in children have demonstrated that an atherogenic pattern of risk factors can start in childhood [24,25]. Such a pattern has been noted in children with PP [24], but that study was limited to girls. Obesity and cardiovascular risk factors track from childhood into adulthood [25,26]. Thompson et al [26] have shown that CVD risk factors are present in overweight children at the age 9 of years, but this has not been demonstrated in lean children with PP. In this study, we demonstrate that both boys and girls with PP have significantly higher total cholesterol (TC) to highdensity lipoprotein (HDL) ratios compared with controls matched for age, sex, ethnicity, and BMI. Furthermore, increased adiposity, serum dehydroepiandrosterone sulfate (DHEAS), and plasma cytokine levels as early markers of metabolic syndrome were found in lean children with PP. These findings are consistent with early onset of metabolic syndrome in children with PP and normal BMI (b19 kg/m2). 2. Methods
stage II or more pubic hair development in either sex, Tanner stage 1 breast in girls, and testicular volume of less than 4 mL in boys) at diagnosis were included in the study. 3.2. Exclusion criteria Before enrollment in the study, all patients with PP underwent an adrenocorticotropic hormone stimulation test to exclude those with late-onset forms of congenital adrenal hyperplasia, and IV leutinizing hormone releasing hormone (LHRH) [28] and radiological investigations, if indicated, to exclude adrenal and ovarian tumors and central precocious puberty. These tests are standard in the evaluation of children with PP and not part of the experimental design for this study. Children with PP secondary to any of the above diagnoses were excluded from the study. 3.3. Final subject group There were 10 PP subjects (ST): 7 girls and 3 boys (mean age ± SD, 7.3 ± 1 years; age range, 5.9-8.98 years). Seven of the total group of 10 children (5 girls and 2 boys [S1]) were separately analyzed because of their normal BMI (b19 kg/m2, ie, b95th percentile at age 7 years) (mean age ± SD, 7.2 ± 1.2 years; age range, 5.9-8.98 years).
4. Controls Prepubertal controls without PP, but matched for age (±2 years), sex, ethnicity, BMI (±2 kg/m2), and maternal history of gestational diabetes were identified for each of the PP subjects. Hence, 3 control groups were matched one to one for each of the subjects: (a) prepubertal girls younger than 6 years if African American, (b) prepubertal girls younger than 8 years if white, and (c) prepubertal boys of either ethnic origin younger than 9 years.
2.1. Subjects: informed consent
4.1. Method of recruitment
All children enrolled in the study and their parents were thoroughly familiarized with the protocols, including any risks of the procedure described below and the potential benefits gained as a result of the study. The study was approved by the Vanderbilt University Institutional Review Board Human Subjects Committee. Participants were permitted to terminate their involvement in the project at any time during their enrollment. Written informed consent was obtained from parents or guardians of all participants, and assent was also obtained from the participant child.
Control children were recruited from the General Pediatric or Adolescent Clinic, local elementary schools, private pediatric practices, or through advertisement in the Vanderbilt University media. 4.2. Exclusion criteria Controls were screened for hyperlipidemia, and those with fasting cholesterol N200 mg/dL and/or triglycerides N150 mg/dL were excluded as controls. 4.3. Final control group
3. Recruitment of subjects 3.1. Eligibility criteria Girls before the age of 6 years, if African American, or before the age of 8 years, if white [27], and boys below the age of 9 years in either ethnicity with isolated PP (Tanner
There were 7 girls and 3 boys in the matched total control group (CT) (mean age ± SD, 7.3 ± 1.6 years; age range, 3.499.25 years), from which 5 girls and 2 boys were selected based on their match to normal-BMI PP group (S1) and called normal-BMI control group (C1) (mean age ± SD, 7.3 ± 2 years; age range, 3.49-9.25 years).
R.P. Mathew et al. / Metabolism Clinical and Experimental 57 (2008) 733–740
735
Table 1 Anthropometric characteristics Parameter Subjects n Age (y) Sex F/M Weight SD Weight percentile Height SD Height percentile Body surface area (m2) Waist-hip ratio BMI (kg/m2) BMI (SD) Systolic BP (mm Hg) Systolic BP corrected for height (mm Hg/m) Diastolic BP (mm Hg) Diastolic BP corrected for height (mm Hg/m) MAP (mm Hg) MAP corrected for height (mm Hg/m)
All a
Normal BMI a
PP (ST) 10 7.3 ± 1.0 7:3 1.2 ± 1.1 78.2 ± 23.8 1.2 ± 0.8 81.9 ± 13.1 1.06 ± 0.2 0.9 ± 0.1 18.8 ± 4.7 1.3 ± 0.8 109.5 ± 9 84.7 ± 8
Controls (CT) 10 7.3 ± 1.6 7:3 0.8 ± 0.9 69.5 ± 19.1 0.6 ± 0.7 61.6 ± 25.8 ⁎ 1.04 ± 0.2 0.9 ± 0.1 19.0 ± 4.8 1.2 ± 0.8 102.0 ± 6.7 80.2 ± 8.8
PP a (S1) 7 7.2 ± 1.2 5:2 0.6 ± 0.7 69.5 ± 23.5 0.9 ± 0.6 76.1 ± 11.2 1 ± 0.1 0.9 ± 0.1 16.3 ± 2 0.9 ± 0.5 110.7 ± 10.3 86.7 ± 7.9
Controls a (C1) 7 7.3 ± 2.0 5:2 0.4 ± 0.3 61.1 ± 13 0.6 ± 0.6 61.3 ± 24 1 ± 0.2 1 ± 0.2 16.4 ± 1.5 0.8 ± 0.4 99.3 ± 2.9 ⁎ 80.6 ± 9.1
59.4 ± 6.3 46 ± 6.3
53.0 ± 5.6 ⁎ 41.8 ± 6.6
61.4 ± 6.5 48 ± 6.2
52.2 ± 1.6 ⁎ 42.5 ± 5.8
79.7 ± 10.4 61.6 ± 8.9
70.9 ± 3.9 ⁎ 55.9 ± 7
83.1 ± 10.8 64.9 ± 8.3
70.7 ± 2.7 ⁎ 57.4 ± 6.8
a Continuous variables are presented as means ± SD. MAP indicates mean arterial pressure. ⁎ P ≤ .05.
All subjects and controls underwent initial clinical evaluation by the principal investigator including pubertal staging according to Marshall and Tanner [29] to confirm that both groups were prepubertal.
based on the method of Teitz [33] (Alpha Wasserman, ACE HDL-C). Atherogenic index was calculated from the following formula: (TC − HDL)/HDL [34]. Adiponectin and C-reactive protein (CRP) levels were determined by a double antibody sandwich enzyme-linked immunosorbent assay method.
4.5. Height and weight measurements
6.1. Dehydroepiandrosterone sulfate
Height was measured 3 times using a Harpenden stadiometer, and the average was taken in deciding the actual height. Weight was measured without shoes using a digital weighing machine once. Weight and height SD scores, BMI [30], surface area, and waist-hip ratio were calculated.
The measurement of DHEAS was accomplished using a kit from Diagnostic Systems Laboratories, Webster, TX, according to the manufacturer's instructions. The sensitivity of the assay was around 1.7 μg/dL, and between-assay coefficient of variation was 10%. There is a 40% crossreactivity with dehydroepiandrosterone.
5. Blood pressure measurements
6.2. Cytokines
Blood pressure (BP) in the recumbent position was determined 3 separate times over a span of 15 minutes using the Dinamap (Critikon Corp, Tampa, FL) and once manually by the nurse coordinator: the average of the above 4 readings was recorded for comparison of systolic, diastolic, and mean arterial BP between PP subjects and their matched controls.
Tumor necrosis factor (TNF)–α, interleukin (IL)-8, IL1B, IL-10, IL-12p70, and IL-6 were assayed by human inflammation cytokine cytometric bead array kit (BD Biosciences Pharmingen, San Diego, CA).
6. Measurement of plasma lipids, adiponectin, and C-reactive protein
The dual-energy x-ray absorptiometry (DEXA) scanner used was a Lunar Prodigy model (software version 8.1 GE Lunar Prodigo Pro, MA). The raw scan data, which include values of tissue and bone, were captured and sent to a computer for quantitative assessment. The term android is used for central obesity, and gynoid is used for obesity of the lower body often seen in women [35]. The radiation exposure for the DEXA was 7 mrem.
4.4. Initial evaluation of the phenotype
Fasting blood samples were collected for the measurements. Determination of cholesterol was based on a modification of the method published by Allain et al [31], that of triglyceride was based on the method of Fossati and Prencipe [32], and that of HDL cholesterol (HDL-C) was
7. Dual-energy x-ray absorptiometry scans
736
R.P. Mathew et al. / Metabolism Clinical and Experimental 57 (2008) 733–740
8. Correction of data for height To control for height, systolic, diastolic, and mean arterial pressures were recalculated for individual patient height (meters) [36]. In addition, serum lipids, DHEAS, biometric measurements, and biochemical markers (cytokines, CRP, and adiponectin) were corrected for height (meters); and significant corrected-for-height values (P ≤ .05) were included in the tables. 8.1. Statistics We assessed pairwise differences between PP subjects and controls using the McNemar test for categorical variables, and a paired t test for continuous variables, when normally distributed. Wilcoxon signed rank test was used when the data were skewed. P values less than .05 or less were considered statistically significant. All tests were 2-tailed. Statistical analyses were performed on a personal computer with the statistical package SPSS for Windows (Version 11.5; SPSS, Chicago, IL) and the statistical language R (www.r-project.org ). 9. Results 9.1. Clinical findings Despite matched BMI in subjects and controls, PP subjects (ST) had a significantly higher height percentile (P = .046) and mean arterial pressure (P = .037) compared with matched controls (CT). The PP subjects with normal BMI (S1) also showed significantly higher systolic BP (P = .028), diastolic BP (P = .028), and mean arterial pressure (P = .018) compared with matched controls (C1) (Table 1). Nevertheless, for both groups, despite the trends, all the above parameters were not significant when corrected for height (ST: mean arterial pressure, 0.13; CT:
Fig. 1. The TC/HDL ratio in controls and subjects.
systolic BP, P = .177; diastolic BP, P = .15; mean arterial pressure, 0.123). 9.2. Biochemical data The PP subjects in either group (ST and S1) had significantly higher DHEAS compared with their respective matched controls (CT and C1, respectively) (P = .043) even after correction for height in both groups (ST vs CT total group, P = .021; ST vs CT normal-BMI group, P = .028) (Table 2), confirming our previously described adrenal source of the androgens [28]. The PP subjects also had higher TC/HDL ratio (p: 0.042) (Fig. 1) and atherogenic index (P = .0.042). The TC/HDL ratio (P = .059) and atherogenic index (p: 0.059) in PP subjects were not significantly elevated when corrected for height. There were nonsignificant trends in increased plasma TC, low-density lipoprotein (LDL), and triglycerides (TG) in PP subjects; and the LDL cholesterol (LDL-C) (milligrams
Table 2 Serum lipid and DHEAS characteristics Parameter Subject n TC (mg/dL) b HDL-C (mg/dL) b LDL-C (mg/dL) b LDL/HDL ratio TC/HDL ratio Atherogenic index TG (mg/dL) c TG/HDL ratio DHEAS (μg/dL) d DHEAS (μg/dL)/height (meter) a
All a
PP (ST) 10 158.7 ± 31.2 47.2 ± 11 94.5 ± 22.6 2.1 ± 0.6 3.53 ± 0.8 2.5 ± 0.8 85.2 ± 44.8 1.69 ± 1.04 80.8 ± 39.6 61.6 ± 31.2
Normal BMI a
Controls (CT) 10 140.1 ± 14.3 46.4 ± 8.3 82.6 ± 10.6 1.8 ± 0.3 3.1 ± 0.4 ⁎ 2.1 ± 0.4 ⁎ 55.9 ± 20.5 1.42 ± 0.61 33.9 ± 21.1 ⁎ 24.9 ± 14.5 ⁎
Continuous variables are presented as means ± SD. Cholesterol in milligrams per deciliter × 0.0259 = millimoles per liter. c Triglycerides in milligrams per deciliter × 0.01 = grams per liter. d Dehydroepiandrosterone sulfate in micrograms per deciliter × 0.026 = micromoles per liter. ⁎ P ≤ .05. b
PP a (S1) 7 147.1 ± 25.7 50.6 ± 11.4 84.7 ± 14.9 2.2 ± 0.5 3.0 ± 0.4 2.0 ± 0.4 62.6 ± 22.7 1.83 ± 1.21 74.8 ± 46.9 57.9 ± 38.0
Controls a (C1) 7 141.1 ± 16.7 49.6 ± 8.9 82.6 ± 9.7 1.9 ± 0.3 2.9 ± 0.3 1.9 ± 0.3 46.1 ± 7.7 1.46 ± 0.54 29.3 ± 16.2 ⁎ 22.0 ± 11.0 ⁎
R.P. Mathew et al. / Metabolism Clinical and Experimental 57 (2008) 733–740
737
Table 3 Variations in extremity and trunk fat and fat distribution by DEXA, and abdominal skin fold thickness Parameter Subject N Left arm fat (g) Total left-sided tissue fat (g) Right arm fat (g) Total right-sided tissue fat (g) Arms fat (g) Fat distribution: android Fat distribution: A/G Trunk fat % Trunk fat (g) Total fat (g) Total body fat distribution % Abdominal skin fold (mm)
All a
PP (ST) 10 518.5 ± 408.5 5808.1 ± 4383.6 509.8 ± 421.7 5720.4 ± 4301.7 1028.6 ± 827.4 37.5 ± 15.2 0.88 ± 0.2 30.6 ± 14.1 5635.7 ± 4914.3 11528.6 ± 8684.4 31.3 ± 12.8 21.6 ± 13.2
Normal BMI a
PP a(S1) 7 283.4 ± 142.5 3334.6 ± 1466.8 272.7 ± 145.6 3292.3 ± 1394.8 556.6 ± 287.2 30.2 ± 11.8 0.8 ± 0.2 23.6 ± 9.8 2850 ± 1307.9 6627 ± 2859.1 25.1 ± 9.1 17.2 ± 10.4
Controls (CT) 10 443.8 ± 383.3 ⁎ 4879.1 ± 4033.8 ⁎ 406.6 ± 430 4681.2 ± 3822.2 ⁎⁎ 904.6 ± 814.7 ⁎ 31.6 ± 15.4 ⁎ 0.76 ± 0.2 ⁎ 25.8 ± 13.9 ⁎ 4158.9 ± 3847.1 ⁎⁎ 9560.4 ± 7855.4 27.5 ± 11.9 14. 9 ± 9.2
Controls a (C1) 7 222.4 ± 82.6 2623.4 ± 779.2 218.9 ± 96.7 2538.1 ± 785.4 441.4 ± 177.7 22.6 ± 5.7 0.65 ± 0.2 ⁎ 17.8 ± 4.8 2014.4 ± 810.1 5161.7 ± 1563.2 20.6 ± 3.5 10.2 ± 3.5
A/G indicates android:gynoid ratio. a Continuous variables are presented as means ± SD. ⁎ P ≤ .05. ⁎⁎ P ≤ .01.
per deciliter) was 94.5 ± 22.6 in PP subjects vs 82.6 ± 10.6 in controls (P = .123) (Table 2). 9.3. Biometric parameters The PP subjects had higher fat (grams) in arms (P = .023), total left side fat tissue (grams) (P = .011), and total right side fat tissue (grams) (P = .004). Fat distribution in PP subjects indicated significantly higher android (P = .047) and android-gynoid ratio (P = .013). Normal-BMI PP children had significantly higher android-gynoid ratio fat distribution compared with their matched controls (P = .037). Trunk fat percentage (P = 0.04) and trunk fat (grams) (P = .007) were significantly elevated in PP children compared with matched controls (Table 3). None of the biometric parameters were significantly elevated when corrected for height (fat [grams] in arms, P = .177; total left side fat tissue [grams], P = .139; total right side fat tissue [grams], P = .459; android fat distribution, P = .207; android-gynoid ratio, P = .175; trunk
fat percentage, P = .193; and trunk fat [grams], P = .328) in PP children compared with matched controls in the total and normal-BMI groups. 9.4. Cytokines The PP subjects had significantly higher TNF-α (P = .038) (Fig. 2) and IL-8 (picograms per milliliter) (P = .05) cytokines compared with matched controls. Normal-BMI PP children also had higher TNF-α (P = .028) (Fig. 2) compared with matched controls. The TNF-α was significantly higher in PP subjects compared with controls in both total and normal-BMI groups when corrected for height (total group, P = .037; normal-BMI group, P = .043) (Table 4). The IL-8 (picograms per milliliter) levels in PP subjects were not statistically different when corrected for height (P = .546). Table 4 Serum cytokines (picograms per milliliter), CRP (nanograms per milliliter), and adiponectin (nanograms per milliliter) Parameter
a
Normal BMI
Subject
PP (ST)
Controls (CT)
PP a (S1)
n TNF-α TNF-α/ height (m) IL-8 IL-1B IL-10 IL-12p70 IL-6 CRP
10 1.8 ± 0.9 1.5 ± 0.6
10 0.8 ± 0.7 ⁎ 0.6 ± 0.5 ⁎
7 1.8 ± 1 1.4 ± 0.7
Controls a (C1) 7 0.6 ± 0.2 ⁎ 0.5 ± 0.5 ⁎
2.6 ± 0.4 5.6 ± 1.2 2.7 ± 1.5 1.2 ± 0.2 2.8 ± 0.7 830.7 ± 923.3 14012.7 ± 8165.8
1.7 ± 1.3 ⁎ 3.8 ± 2.4 2.2 ± 1.6 1 ± 0.8 3.0 ± 1.2 1332.7 ± 1529.8 14068.97 ± 8339.2
2.6 ± 0.4 5.6 ± 1.3 2.5 ± 1.8 1.2 ± 0.2 2.8 ± 0.8 640.5 ± 923.4 16256.5 ± 8965.6
1.4 ± 0.5 3.8 ± 2.9 1.8 ± 1.7 0.9 ± 1 2.8 ± 1.4 1332.7 ± 1529.8 17329.9 ± 7567.1
Adiponectin
Fig. 2. Differences in cytokine TNF-α in controls and subjects.
All a
a Continuous variables are presented as means ± SD. ⁎ P ≤ .05.
738
R.P. Mathew et al. / Metabolism Clinical and Experimental 57 (2008) 733–740
10. Discussion After having excluded tumors and central or peripheral activity of the gonads, we confirmed in the study that the principal cause of PP was adrenal hyperandrogenism, as implied in the term premature adrenarche (PA). We found that PP subjects had a significantly higher ratio of cholesterol to HDL; greater peripheral, truncal, and total body fat; and higher plasma levels of cytokines TNF-α and IL-8 than BMImatched controls. Of these, only TNF-α was significantly and consistently elevated in total and normal-BMI groups of PP children when compared with matched controls and when corrected for height in both groups. These changes in serum cytokines in PP children with normal BMI have not been previously reported. Premature adrenarche was previously considered to be a benign condition; but more recently, it has been associated with hyperinsulinemia, dyslipidemia, and hypertension [37]. Elevated serum triglycerides and an increased ratio of LDLC to HDL-C have been found in girls with PA [24]. The dyslipidemia of insulin resistance has been proposed to be a precursor to the development of atherosclerosis [38]. Ibanez et al [24] showed that girls with PA have higher triglyceride, cholesterol, LDL-C, and VLDL levels than their healthy peers through the prepubertal and pubertal periods. Hyperinsulinemia and insulin resistance, both independently and together, have been hypothesized to result in dyslipidemia in subjects with both normal and impaired glucose tolerance [39,40]. Hyperinsulinemia enhances hepatic VLDL synthesis [38] and could directly contribute to the increased triglyceride and cholesterol values in these subjects. Laws and Reaven [41] showed that a high triglyceride and a low HDL-C concentration are strong indicators of insulin resistance [42]. In our study, TC/HDL and atherogenic index were significantly elevated in the total group of PP subjects. Nevertheless, the above findings for both groups did not reach significance when corrected for their respective heights. Guven et al [34], in a study of Turkish girls, demonstrated that those with PA have higher BMI, elevated LDL-C and VLDL-C levels, hyperinsulinemia, and higher BP than their healthy peers. However, cholesterol differences have not been previously demonstrated in a mixed group of children with PP matched with controls for BMI, in addition to age, ethnicity, and sex. There have been major advances in the noninvasive assessment of body composition. In this study, we used supine DEXA to assess body composition [43]. Along with underwater weighing and isotope dilution, DEXA is considered the most accurate method for measurement of body composition [44]. To provide consistency in body composition measurements, all PP subjects and controls in our study underwent DEXA scan measurement of body composition in supine position. We have shown higher android-gynoid ratio fat distribution in PP children with normal BMI compared with matched controls. This is likely
to be directly related to increased DHEAS levels in PP. Sex steroids play important roles in determining body fat mass and its distribution [45] including body fat patterning [46-48]. It has also been suggested that intraadipose sex steroid metabolism is also a determinant of android vs gynoid patterns of body fat [45]. These differences in body and extremity fat composition in children with PP and normal BMI indicate the influence of fat metabolism in PP subjects even when lipid profiles and BMI are in the reference range and comparable with matched controls. Adipose tissues secrete a variety of molecules and adipocytokines, including TNF-α and IL-8 [49]. Previous studies have shown that elevated serum levels of CRP, IL-6, and TNF-α precede type 2 diabetes mellitus [50] and coronary heart disease [51]. In our study, we found marked elevations in TNF-α and IL-8 initially and elevation of TNFα alone for both groups when corrected for their individual height. Elevation of TNF-α in PP children regardless of BMI suggests a possible direct correlation to later development of essential hypertension in metabolic syndrome. Further studies are needed to confirm this hypothesis. In addition, TNF-α was consistently elevated in obese and nonobese children with PP and, if confirmed in bigger studies, may be an early predictor of metabolic syndrome at PP diagnosis. In children with PP, our results support the contention that the cluster of risk factors for CVD starts in childhood [26,52]. The increased insulin concentrations concurrent with altered lipid profile may initiate this process in the prepubertal period [53]. Based on our data, the development of PP regardless of BMI should be interpreted as indicative of insulin resistance, a proinflammatory environment, and a risk for development of metabolic syndrome. Metformin has been advocated for use in children with insulin-resistant syndromes and is considered safe [54,55]. Tumor necrosis factor–α down-regulation may be accomplished by an antibody to TNF-α (as infliximab in rheumatoid arthritis) [56], although the sole use of TNF-α antibody failed to reverse insulin resistance in humans. Our data suggest that additional therapeutic targets may be identified for the inflammatory cytokines that are elevated in children with PP. In conclusion, children with PP, even those with normal BMI, show early signs of the metabolic syndrome and should be targeted with early lifestyle interventions. Our study suggests that TNF-α levels are markers of metabolic syndrome at PP diagnosis. Acknowledgment The study was supported by Independent Clinical Research Grant from Pfizer and by the General Clinical Research Center (GCRC), Vanderbilt University, Nashville, TN, and GCRC Immunology Core through GCRC Grant MO1 RR00095 from the National Center for Research Resources/National Institutes of Health. We are indebted to Dr Marta Hernanz-Schulman for her critical review of the manuscript.
R.P. Mathew et al. / Metabolism Clinical and Experimental 57 (2008) 733–740
References [1] Saenger P, Reiter E. Premature adrenarche: a normal variant of puberty. J Clin Endocrinol Metab 1992;74:236-8. [2] Reiter E. Premature adrenarche. Endocrinologist 1997;7:85. [3] Ibanez L, de Zegher F, Potau N. Premature pubarche, ovarian hyperinsulinism, hyperandrogenism and the polycystic ovary syndrome: from a complex constellation to a simple sequence of prenatal onset. J Endocrinol Invest 1998;21:558-66. [4] Rittmaster R, Deshwal N, Lehman L. The role of adrenal hyperandrogenism, insulin resistance, and obesity in the pathogenesis of polycystic ovary syndrome. J Clin Endocrinol Metab 1993;76:1295-9. [5] DiMartino-Nardi J. Premature adrenarche: findings in prepubertal African American and Caribbean-Hispanic girls. Acta Paediatr Suppl 1999;433:67-72. [6] Eckel R, Grundy S, Zimmet P. The metabolic syndrome. Lancet 2005; 365:1415-28. [7] Jones K. The dilemma of the metabolic syndrome in children and adolescents: disease or distraction? Pediatr Diabetes 2006;7:311-21. [8] Adults., E.P.o.D.E.a.T.o.H.B.C.i., Executive summary of the third report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). JAMA 2001;285: 2486-97. [9] Cook S, et al. Prevalence of a metabolic syndrome phenotype in adolescents: findings from the third national health and nutrition examination survey. 1988-1994. Arch Pediatr Adolesc Med 2003;157: 821-7. [10] De Ferranti S, et al. Prevalence of metabolic syndrome in American adolescents. Circulation 2004;110:2494-7. [11] Goodman E, et al. Contrasting prevalence of and demographic disparities in the World Health Organization and National Cholesterol Education Program Adult Treatment Panel III definitions of metabolic syndrome among adolescents. J Pediatr 2004;145:445-51. [12] Weiss R, et al. Obesity and metabolic syndrome in children and adolescents. N Engl J Med 2004;350:2362-74. [13] Alberti K, Zimmet P, Shaw J. The metabolic syndrome—a new worldwide definition. Lancet 2005;366:1059-62. [14] Grundy S. Metabolic syndrome: a multiplex of cardiovascular risk factor. J Clin Endocrinol Metab 2007;92(2):399-404. [15] Ferrannini E. Metabolic syndrome: solution in search of a problem. J Clin Endocrinol Metab 2007;92(2):396-8. [16] Denburg M, et al. Insulin sensitivity and the insulin-like growth factor system in prepubertal boys with premature pubarche. J Clin Endocrinol Metab 2002;87:5604-9. [17] Ibanez L, et al. Fat distribution in non-obese girls with and without precocious pubarche: central adiposity related to insulinaemia and androgenaemia from prepuberty to postmenarche. Clin Endocrinol 2003;58:372-9. [18] Bouchard C, Bray GA, Hubbard VS. Basic and clinical aspects of regional fat distribution. Am J Clin Nutr 1990;52:946-50. [19] Daniels S, et al. Association of body fat distribution and cardiovascular risk factors in children and adolescents. Circulation 1999;99:541-5. [20] Walton C, et al. Body fat distribution, rather than overall adiposity, influences serum lipids and lipoproteins in healthy men independent of age. Am J Med 1995;99:459-64. [21] Ashwell M. Obesity in men and women. Int J Obes Relat Metab Disord 1994;18:S1-S7. [22] Smoak C, et al. Relation of obesity to clustering of cardiovascular disease risk factors in children and young adults. The Bogalusa Heart Study. Am J Epidemiol 1987;125:364-72. [23] Sardinha L, et al. Subcutaneous central fat is associated with cardiovascular risk factors in men independently of total fatness and fitness. Metabolism 2000;49:1379-85. [24] Ibanez L, et al. Hyperinsulinemia, dyslipidemia and cardiovascular risk in girls with a history of premature pubarche. Diabetologica 1998;41: 1057-63.
739
[25] Freedman D, et al. Inter-relationships among childhood BMI, childhood height, and adult obesity: the Bogalusa Heart Study. Int J Obes 2004;28:10-6. [26] Thompson D, et al. Childhood overweight and cardiovascular disease risk factors: the National Heart, Lung, Blood Institute Growth and Health Study. J Pediatr 2007;150:18-25. [27] Herman-Giddens M, et al. Secondary sexual characteristics and menses in young girls seen in office practice: a study from the pediatric research in office settings network. Pediatrics 1997;99:505-12. [28] Mathew R, et al. Premature pubarche in girls is associated with functional adrenal, but not ovarian, hyperandrogenism. J Pediatr 2002; 102:91-8. [29] Marshall W, Tanner J. Variations in pattern of pubertal changes in girls. Arch Dis Child 1969;44:291-303. [30] Grinstein G, et al. The relationship between birth weight (BW), body mass index (BMI) and insulin sensitivity (SI) in prepubertal Caribbean Hispanic (CH) and black African American (BAA) girls with premature pubarche.Pediatr Res 1999;45:89 [abstract]. [31] Allain C, et al. Enzymatic determination of total serum cholesterol. Clin Chem 1974;20:470-5. [32] Fossati P, Prencipe L. Serum triglycerides determined calorimetrically with an enzyme that produces hydrogen peroxide. Clin Chem 1982;28: 2077-80. [33] Teitz N. Clincal guide to laboratory tests. Vol. 3rd ed. Philadelphia, PA: Saunders; 1995. [34] Guven A, Cinaz P, Bideci A. Is premature pubarche a risk factor of atherogenesis? Pediatr Int 2005;47:20-5. [35] Flier J, Foster D. Eating disorders: obesity, anorexia nervosa, and bulimia nervosa in William textbook of endocrinology. In: Flier J, editor. Eating disorders: obesity, anorexia nervosa, and bulimia nervosa. 9th ed. Philadelphia, PA: Saunders; 1998. [36] Michels K, Greenland S, Rosner B. Does body mass index adequately capture the relation of body composition and body size to health outcomes? Am J Epidemiol 1998;147:167-72. [37] Saenger P, DiMartino-Nardi J. Premature adrenarche. J Endocrinol Invest 2001;24:724-33. [38] Orchard T, et al. Plasma insulin and lipoprotein concentrations: and atherogenic association. J Epidemiol 1983;118:326-37. [39] Abbott W, Lillioja S, Young A. Relationship between plasma lipoprotein concentrations and insulin action in obese hyperinsulinemic population. Diabetes 1987;36:897-904. [40] Laakso M, Sarlund H, Mykkanen L. Insulin resistance is associated with lipid and lipoprotein abnormalities in subjects with varying degrees of glucose intolerance. Atherosclerosis 1990; 10:223-31. [41] Laws A, Reaven G. Evidence for an independent relationship between insulin resistance and fasting plasma HDL-cholesterol, triglyceride and insulin concentrations. J Inern Med 1992;231:25-30. [42] McLaughlin T, et al. Use of metabolic markers to identify overweight individuals who are insulin resistant. Ann Int Med 2003;139:802-9. [43] Lambrinoudaki I, et al. Body composition assessment by dual-energy X-ray absorptiometry: comparison of prone and supine measurements. Metabolism 1998;47:1379-82. [44] Newton R, et al. Percent body fat measured by BIA and DEXA in obese, African-American adolescent girls. Int J of Obesity 2005;29: 594-602. [45] Wake D, et al. Intra-adipose sex steroid metabolism and body fat distribution in idiopathic human obesity. Clin Endocrinol 2007;66: 440-6. [46] Marin P, Bjorntorp P. Endocrine-metabolic patterns and adipose tissue distribution. Horm Res 39:81-85 1993;39:81-5. [47] Rosenbaum M, Leibel R. Role of gonadal steroids in the sexual dimorphisms in body composition and circulating concentrations of leptin. J Clin Endocrinol Metab 1999;77:1041-5. [48] Klein K, et al. Estrogen levels in childhood determined by an ultrasensitive recombinant cell bioassay. J Clin Invest 1994;94: 2475-80.
740
R.P. Mathew et al. / Metabolism Clinical and Experimental 57 (2008) 733–740
[49] Salmenniemi U, et al. Multiple abnormalities in glucose and energy metabolism and coordinated changes in levels of adiponectin, cytokines, and adhesion molecules in subjects with metabolic syndrome. Circulation 2004;110:3842-8. [50] Schimidt M, et al. Markers of inflammation and prediction of diabetes mellitus in adults (Atherosclerosis Risk in Communities Study): a cohort study. Lancet 1999;353:1649-52. [51] Rifai N, Ridker P. Inflammatory markers and coronary heart disease. Curr Opin Lipidol 2002;13:383-9. [52] Arslanian S, Suprasongsin C. Insulin sensitivity, lipids, and body composition in childhood. Is “syndrome X” present? J Clin Endocrinol Metab 1996;81:1058-62.
[53] Ronnemaa T, et al. Serum insulin and other cardiovascular risk indicators in children, adolescents and young adults. Ann Med 1991; 23:63-72. [54] Freemark M, Bursey D. The effect of metformin on body mass index and glucose tolerance in obese adolescents with fasting hyperinsulinemia and a family history of type 2 diabetes. Pediatrics 2001;107:E55. [55] Kay J, et al. Beneficial effects of metformin in normoglycemic morbidly obese adolescents. Metabolism 2001;50:1457-61. [56] Bondeson J, Maini R. Tumor necrosis factor as a therapeutic target in rheumatoid arthritis and other chronic inflammatory diseases: the clinical experience with infliximab (Remicade). Int J Clin Pract 2001; 55:211-6.